Trajectory-Learned Actuator-Aware Funnel Network / bench_report.json
Beats tuned baseline
1{
2 "bench_version": 1,
3 "track": "dynamics",
4 "model": "rnn_small",
5 "metric_direction": "lower is better",
6 "n_seeds": 8,
7 "baseline": {
8 "best_cfg": {
9 "lr": 0.01
10 },
11 "sweep": [
12 {
13 "cfg": {
14 "lr": 0.001
15 },
16 "mean": 0.005253629118669778
17 },
18 {
19 "cfg": {
20 "lr": 0.003
21 },
22 "mean": 0.001200278929900378
23 },
24 {
25 "cfg": {
26 "lr": 0.01
27 },
28 "mean": 0.0006579469481948763
29 }
30 ],
31 "full": {
32 "mean": 0.0006267142089200206,
33 "std": 0.0001539544331865635,
34 "per_seed": [
35 0.0006342190317809582,
36 0.0005908075836487114,
37 0.0004556644707918167,
38 0.0009510967065580189,
39 0.000620594946667552,
40 0.00040738406823948026,
41 0.0006597545580007136,
42 0.0006941923056729138
43 ],
44 "n": 8
45 }
46 },
47 "idea": {
48 "mean": 2.24928216994158e-05,
49 "std": 4.166535651612339e-06,
50 "per_seed": [
51 2.433888403174933e-05,
52 2.139313619409222e-05,
53 2.2563392121810466e-05,
54 2.3747454179101624e-05,
55 2.724600926740095e-05,
56 2.3439619326381944e-05,
57 2.4891782231861725e-05,
58 1.2322296242928132e-05
59 ],
60 "n": 8,
61 "best_cfg": {
62 "lr": 0.01
63 },
64 "sweep": [
65 {
66 "cfg": {
67 "lr": 0.001
68 },
69 "mean": 3.181707052135607e-05
70 },
71 {
72 "cfg": {
73 "lr": 0.003
74 },
75 "mean": 2.6148483357246732e-05
76 },
77 {
78 "cfg": {
79 "lr": 0.01
80 },
81 "mean": 2.301071663168841e-05
82 }
83 ]
84 },
85 "comparison": {
86 "delta_mean": -0.0006042213872206048,
87 "idea_wins": 8,
88 "n_pairs": 8,
89 "per_seed_diffs": [
90 -0.0006098801477492088,
91 -0.0005694144474546192,
92 -0.00043310107867000625,
93 -0.0009273492523789173,
94 -0.000593348937400151,
95 -0.0003839444489130983,
96 -0.0006348627757688519,
97 -0.0006818700094299857
98 ],
99 "p_value": 0.0081,
100 "mde": 0.0001379547529361776,
101 "mde_rel_pct": 22.012386343355264,
102 "verdict": "idea better (significant)",
103 "system_worked": true
104 },
105 "mechanism_signature": {
106 "custom_track": null,
107 "predicted_kmax": 1.5,
108 "observed_authority_mean": 0.08706140890717506,
109 "requested_authority_mean": 0.08706140890717506,
110 "authority_ratio_mean": 0.99999998849756,
111 "funnel_violation_rate_mean": 0.0,
112 "confirmed": true
113 }
114}